What this category's data looks like
Graduate training process management involves materials such as training plans, proposal reports, and mid-term assessment forms. They come from submissions by departments and supervisors and are archived by the graduate school. The update cadence advances with semesters and training stages. Field characteristics include student name, student ID, course name, credits, date, and signature fields. When the materials enter the document parsing and chunking function, the training plan carries the course and credit structure, the proposal report carries the topic selection and time nodes, and the mid-term assessment form records the stage conclusion and signature receipt. Together, the three form the original source text for business basis queries. After parsing, each becomes a visible data row and is bound to a unique data identifier, for subsequent retrieval and for referring back to the original text when replying to parents or writing plans.
Configuration and implementation checks
This step is executed by parsing and training tasks (server-side tasks), outputting data rows and index status (parsed, indexing, indexed). The basis entries that can refer back to the original text are completed next by retrieval and subsequent processing.
| Object | Check | Verification action |
|---|---|---|
| Training plan | Course name, credits, date, and signature fields must correspond item by item with the proposal report and mid-term assessment form | Check whether the name, quantity, and date of the same record are consistent across the three materials, check whether signatures and receipts are complete, and confirm that the data row status is indexed |
| Proposal report | Topic name, proposal date, and supervisor signature must match the training plan and mid-term assessment form | Separately check the identity identifiers of the file record and the data row, review the chunking and training settings corresponding to chunkSize, chunkSplitter, and trainingType, and confirm that the version is current |
| Mid-term assessment form | Stage conclusion, assessment date, and signature receipt must be consistent with the training plan and proposal report | Check whether the data rows covered by the rebuild task contain only data with completed indexing, confirm that data pending indexing did not participate in the rebuild, and recognize only parsed, indexing, and indexed as status values |
| Data row | Is in one of the three states and is consistent with the interface display | Open the data row to check the status, confirm it is consistent with the interface display, and failed training data can be directly edited or deleted |
Three common errors
- Symptom: The name, quantity, and date of the same record do not match across two materials. Cause: The training plan, proposal report, and mid-term assessment form are submitted separately, fields are not cross-checked, and after parsing each becomes its own data row.
- Symptom: The version is not synchronized, and the old training plan is still hit by retrieval. Cause: Rebuilding after changing the model or synonyms only covers data with completed indexing, and data pending indexing does not participate in the rebuild chain.
- Symptom: Signatures or receipts are missing, and the basis entry cannot refer back to the original text. Cause: The signature field in the scanned material is blank. Although the data row has been stored after parsing and chunking, the missing item in the original text causes the reference back to fail.
How to verify the setup
Acceptance for this feature: upload one material, observe the data row status transitions and the coverage scope of the rebuild task, confirm that parsing and chunking are immediately stored as visible data rows after parsing is completed and bound to a unique data identifier, that the index transitions through the three states parsed, indexing, and indexed, that rebuilding targets only data with completed indexing, and check that the chunkSize, chunkSplitter, and trainingType settings are consistent with the interface display.
Business test: ask a cross-material basis question, see whether the answer gives a referenceable source, and check that the tokens of the quoted text and the number of array entries are recorded separately, without comparing the number of quoted entries as equal to the quoted content budget.
- Open the data row to check the status, confirm it is in one of parsed, indexing, or indexed and consistent with the interface display, and directly edit or delete failed training data.
- Separately check the identity identifiers of the file record and the data row, check whether the data rows covered by the rebuild task contain only data with completed indexing, and confirm that data pending indexing did not participate in retrieval and rebuilding.